diff options
| author | Chen Zheng <czhengsz@cn.ibm.com> | 2020-06-14 21:34:52 -0400 |
|---|---|---|
| committer | Chen Zheng <czhengsz@cn.ibm.com> | 2020-06-15 00:00:04 -0400 |
| commit | bd7096b977e11293b5a7406c4846ac805197fca5 (patch) | |
| tree | d99d4e8a1de474a832b833e55efd257821a18497 /llvm/lib/CodeGen/SafeStack.cpp | |
| parent | b559535a3a21fa1f3ee20fbcd4bad4bc840d0a49 (diff) | |
| download | llvm-bd7096b977e11293b5a7406c4846ac805197fca5.zip llvm-bd7096b977e11293b5a7406c4846ac805197fca5.tar.gz llvm-bd7096b977e11293b5a7406c4846ac805197fca5.tar.bz2 | |
[PowerPC] fma chain break to expose more ILP
This patch tries to reassociate two patterns related to FMA to expose
more ILP on PowerPC.
// Pattern 1:
// A = FADD X, Y (Leaf)
// B = FMA A, M21, M22 (Prev)
// C = FMA B, M31, M32 (Root)
// -->
// A = FMA X, M21, M22
// B = FMA Y, M31, M32
// C = FADD A, B
// Pattern 2:
// A = FMA X, M11, M12 (Leaf)
// B = FMA A, M21, M22 (Prev)
// C = FMA B, M31, M32 (Root)
// -->
// A = FMUL M11, M12
// B = FMA X, M21, M22
// D = FMA A, M31, M32
// C = FADD B, D
Reviewed By: jsji
Differential Revision: https://reviews.llvm.org/D80175
Diffstat (limited to 'llvm/lib/CodeGen/SafeStack.cpp')
0 files changed, 0 insertions, 0 deletions